The present invention relates to methods of controlling the output of a solid state laser, and more particularly to a method of reducing transient thermal crosstalk in a monolithic array of individually modulated lasers by providing a loss region in each laser whose loss may be modulated by selective application of a low voltage.
Solid state lasers, also referred to as semiconductor lasers or laser diodes, are well known in the art. These devices are based on the p-n junction from semiconductors, and quantum electronics from lasers. The devices generally consist of a layered semiconductor structure having one or more active layers bounded at their ends by cleaved facets which act as mirrors. An optical resonator, or so-called Fabry-Perot cavity is thereby formed. An electrical potential is applied across the one or more active layers. The voltage drives either holes or electrons or both across the p-n junction (i.e., they are "injected"), and when these carriers recombine they emit light. Optical feedback is provided by the cleaved facets to allow "stimulation" of the recombination to provided coherent emission.
This type of laser is employed in communication systems, laser xerography, and other applications where the device's small size, low operating current, and other characteristics are beneficial. A presently active area in research and development for these applications is the integration of a number of solid state lasers or solid state lasers and other devices onto a single substrate. For example, monolithic arrays of independently addressable solid state lasers (referred to herein as "multilaser arrays") are the optical sources of choice for high-speed, high-resolution laser xerographic printers.
In virtually all of the applications of these lasers it is necessary to modulate the output of the laser. Where a number of solid state lasers are integrated onto a single substrate, it is almost always necessary to modulate the output of each laser independently. In some applications this modulation is at a very high frequency, in others it is at a low frequency, and in still others this frequency will vary. Since lasing depends on a current flowing into the active layer, one obvious way to modulate the light output of a laser is to modulate the driving current. In fact, varying the driving current is presently the most common and conventional way in which a laser's output is modulated. However, this method of modulation has a number of distinct drawbacks and disadvantages, among which is transient heating of the chip on which the laser or lasers are formed.
The causes of heating of a laser structure are well reported in the art. For example, see M. Ito and T. Kimura, "Stationary and Transient Thermal Properties of Semiconductor Laser Diodes," IEEE Journ. of Quant. Electronics, vol. QE-17, pp. 787-795, May 1981. Energy is introduced into the active region, by injecting electrons into the conduction band and/or holes into the valence band. A certain number of electrons fall back down into the valence band through nonradiative processes and release their energy in the form of heat. This may be quantified by first assuming that driving current is modulated between the laser's threshold current, I.sub.th, and the operating current, I.sub.0, at which the operating power P.sub.0 is produced. The heat generated by such modulation is then given by EQU .DELTA.Q=[I.sub.0 V.sub.0 -P.sub.0 -P.sub.sp0 ]-[I.sub.th V.sub.th -P.sub.spth ], (1)
where V.sub.0 is the laser voltage at I.sub.0, V.sub.th is the laser voltage at I.sub.th, and P.sub.sp0 and P.sub.spth are the spontaneous emission powers at I.sub.0 and I.sub.th, respectively. Since the laser's junction voltage saturates at threshold, EQU V.sub.0 =V.sub.th +(I.sub.0 -I.sub.th)R.sub.s ( 2)
where R.sub.s is the series resistance of the laser. Also, since the spontaneous emission power saturates at threshold, P.sub.sp0 =P.sub.spth, so that equation (1) can be rewritten as EQU .DELTA.Q=P.sub.0 [V.sub.th +I.sub.0 R.sub.s -.eta.]/.eta. (3)
where P.sub.0 =.eta.(I.sub.0 -I.sub.th).
Transient heating, or heating which changes in time, results in a number of deleterious effects. For example, since a laser's output power is temperature dependent, temperature variations can result in power output "droop." That is, at the time t=0 that the current applied to the laser reaches I.sub.0, the laser will be at a first temperature, and its output power will have a first value, say P.sub.1 as shown in FIG. 1. However, as the applied driving current increases beyond threshold to the operating current, the temperature of the laser increases. This increase occurs much more slowly than the time it takes the laser to begin to lase in response to the driving current, so that this temperature change occurs at a constant driving current. This heating results in a shift in the output characteristics of the laser to that shown for the time t=.infin. of FIG. 1. This will cause the laser to have an output power P.sub.2 which is less that P.sub.1, resulting in output power droop. Thus, the plot of output power versus time of FIG. 2 shows an initial power of P.sub.1 at turn-on, "drooping" to P.sub.2 as the device is operated.
In monolithic multilaser array applications, the thermal variations may lead to emitter crosstalk. That is, when a first laser is modulated the local temperature will vary. Due to the ability and desire to space the individual lasers very close together, this local temperature variation will be transmitted through the chip, causing or accentuating temperature variations in the neighboring lasers on the chip. This renders the power output of the neighboring lasers unpredictable and erratic. Thus, there is a need to minimize the effect modulating one laser may have on the operation of the other lasers.
A related consequence of transient heating of the chip is pattern effects in the output pulse stream. Assume that the laser is modulated at varying frequencies. The time between successive pulses of the laser will then vary. If the time between successive pulses is large, the device will be given sufficient time to cool so that the application of the driving current has a large temperature effect (i.e., a large droop in output power will occur at turn-on). The shorter the time between pulses, the less time the device has to cool between one pulse and the next. Since the power output determines the output energy profile, the amount of droop will determine the amount of variation in the energy in each pulse, which will in turn have a direct effect on the stability of the laser's output. For example, in laser xerographic applications, varying the energy in each output pulse results in varying exposure of the photoreceptor, which translates directly into varying spot size. Thus, a more constant operating temperature of the device will resulting in a more constant output power, a more uniform output energy profile, and ultimately a more uniform spot pattern for the output pulse stream. Again, where two or more lasers are operated on the same chip, transient thermal cross talk will result and the output pattern will be unpredictable and erratic.
Another related consequence of transient heating of a laser is wavelength variation during a pulse and over long streams of pulses. Essentially, the operating wavelength of a solid state laser is dependent, inter alia, on the temperature of operation. If the temperature varies, the wavelength of operation will vary. The effect of this variation of wavelength, for example in the laser xerography application, is to vary the energy that can be written onto the photoreceptor. This also can translate directly into variations in the spot pattern on the photoreceptor. Thus, in order to maintain the output of the various lasers on the chip at the desired lasing wavelength, it is important to minimize the thermal crosstalk between them.
As relationship (3) demonstrates, the temperature change is, in part, a function of the difference between I.sub.th and I.sub.0. Methods of operation have been developed, however, which result in smaller differences between I.sub.th and I.sub.0 than others. Certain of these methods employ a laser structure having two coaxial, independently addressable regions (hereinafter an "amplifier region" and a "modulator region"), each with its own p-n junction, formed between two cleaved facets on a single substrate. Such a structure is described in detail in U.S. Pat. No. 4,802,182, dated Jan. 31, 1989, to Thornton et al., which is incorporated by reference herein. This reference provides several modes of operation. One mode, called the "electro-absorption mode", is one in which the amplifier region is sufficiently forward biased to cause stimulated emission, and a negligible minimal forward bias current is applied to the modulator region as the voltage to the modulator region is varied. In this mode, the modulation is controlled by electro-absorption in the modulation region, wherein as the voltage is increased (from negative to positive), the modulation region becomes more transparent to and less absorbent of the emission from the amplifier region. Another mode, called the "gain modulation mode", also has the amplifier region sufficiently forward biased to cause stimulated emission, but has the DC voltage on the modulator region generally constant while the current is varied. In this mode, modulation of the output will be a result of changes in the emission gain in the modulator region due to the changes in carrier density in that region.
The modulation of a laser by varying the loss in the modulator region is generically referred to as Q-switching. Another approach to modulating a laser based on Q-switching is discussed in Kressel and Butler, Semiconductor Lasers and Heterojunction LEDs, Academic Press, 1977, p. 574. A two region laser is described therein, in which, as with the above references, the amplifier region is strongly pumped with current to serve as a light emitting region, and the modulator region is pumped with a lower current level than the amplifier region to effect high frequency modulation. The operation described in this reference, however, provides self-sustaining modulation. As described in the reference, when a photon is emitted by the first region and absorbed by the second region an electron-hole pair is formed. A following photon with a sufficiently high energy which enters the second region will cause stimulated emission, which reduces the carrier population in the second region for the next incident photon. Thus, if the device parameters and operating conditions are properly selected, self-sustaining periodic modulation occurs.
Another Q-switched method is described in U.S. Pat. No. 3,768,037, dated Oct. 23, 1973, to Migitaka et al. In the method described therein, a laser is provided having a structure similar to the above-mentioned two-region devices, with one region being defined by either a "U" shaped contact or an "L" shaped contact. An amplifier region is sufficiently biased to result in stimulated emission, and a varying reverse bias is applied to a modulation region such that a depletion region is formed therein. Since threshold current varies directly as a function of the loss in the resonator cavity of the laser, the output of the laser may be modulated by varying the loss which may be accomplished by varying the reverse bias on the modulator region.
Each of these Q-switching schemes requires a laser whose structure includes separate amplifier and modulator regions. This geometry is accepted in single laser applications such as that of the aforementioned references. However, Q-switching has not been employed in arrays of lasers because to do so would require the introduction of a modulator region in the device geometry, which increases the loss of the resonator. This increase in loss results in an increase in the threshold current for the laser. In the commonly employed current modulated lasers this increase in threshold means an increase in operating temperature. Since this implies an increase in the temperature differential between the ON and OFF states of the laser, an increase in threshold is expected to result in an increase in thermal cross talk. Thus, not only have multilaser array applications been unable to avail themselves of Q-switching techniques, there has been great effort made to minimize I.sub.th.
Furthermore, this desire to keep the temperature difference between the ON and OFF states to a minimum also leads to operating a laser at low power levels, since the higher the output power level is, the higher the input current is, and the greater the difference between I.sub.th and I.sub.0. This is disadvantageous since it puts great demands on the equipment designed to receive the laser's output.
These and other problems are addressed by various aspects of the present invention, which will be summarized then described in detail below.